The USADA had George Hincapie -- the guy who was Armstrong's domestique through all seven Tour wins -- willing to testify against him. That means a lot. In any event, Armstrong's decision is probably more due to the real Feds now wanting to get a piece of him due to mounting evidence that he was essentially the ringleader of a distribution operation.
The kind of numbers that Tour riders were putting up until just a few years ago are now believed to be physiologically impossible to do without doping. There's a legitimate reason cycling has, quite literally, slowed down since the '90s -- people are climbing Alpe d'Huez about 3-4 minutes slower than a decade ago, and that's an absolute eternity in cycling (and just over one climb) ... it's not an illogical theory that this is due to the sport slowly being cleaned up.
USADA claims it has 38 tests from 2009 - 2010 that show evidence of blood doping, plus the 2001 EPO test. I don't see a problem, here -- Armstrong shouldn't get a free pass just because he got millions of people to wear his bracelets, and with him being such a stubborn guy, I don't think he'd make this decision without believing that they finally had him dead to rights.
And we are two thirds into 2012. If the tests returned positive results in 2001 why wasn't something done about it then? If the tests returned positive results in 2009-2010 why wasn't something done then?
... because testing methods don't remain static over time? Here's a really great read.
Some cancers - including the type Lance Armstrong had - cause enormously elevated levels of human chorionic gonadotropin hormone (hCG), a naturally occuring hormone in the body, but at low levels in males. Now, there are rules for the amount of HCG permitted in an athlete, because it offers a competitive advantage - not enough to overcome the deficiencies cancers cause, but a good advantage in a healthy human being, because it produces testosterone. An athlete is often considered to have failed a drug test if the urinary T/E ratio is greater than 6. So the UCI would have been testing for it.
But Armstrong never produced a positive sample. Compare that with Jake Gibb whose life, it could be argued, was saved by USADA’s testing, when it detected those enormously elevated levels in an anti-doping test, and advised him to see a doctor. That ultimately led to the discovery of testicular cancer, and Gibb recovered. Lance Armstrong wasn’t so lucky - so we can assume one of two things. Either the UCI’s anti-doping measures were woefully below standard, and didn’t detect Armstrong’s elevated levels of hCG, allowing his cancer to worsen while competing, or the UCI’s anti-doping discovered Armstrong’s elevated levels and didn’t report them. Either way, it’s a massive condemnation in the UCI’s ability to validate itself as a serious entity in drug testing. At best it’s woefully ineffective, at worst it’s simply corrupt.
The other, seemingly more likely, answer is that Armstrong was using a masking agent.